Refugees in the Netherlands: Victims or Intruders?
The power of images in media coverage of the refugee crisis
Faten Bushehri
Student number: 11081791
Master’s Thesis
Graduate School of Communication
Master’s programme Communication Science
University of Amsterdam
Supervisor: Jonas Lefevere
Date of submission: 27/05/2016
Abstract
While framing and framing effects theories of text media coverage are extensively studied, visual framing effects often receive less scholarly attention. This article uses the refugee crisis as a case study to test how different visual frames of refugees affect people’s emotions and attitudes towards refugees in the Netherlands. Furthermore, I test whether the attitudinal effect of visual frames is mediated by emotions. In an experimental design, participants were exposed to two visual frames that portray refugees either as victims or intruders, and then measured participants’ level of, empathy, hope, anger and fear as well as their attitude towards welcoming refugees into the Netherlands. Findings show that participants’ feelings of hope and empathy are significantly higher in the victim frame, while there are no significant differences in the emotions of fear and anger between the victim and the intruder frame. Results also show that visual frames have no significant effects on participants’ tolerance towards refugees coming to their country. The study found that empathy mediated the effect of the victim frame on participants’ tolerance levels, while anger, fear and hope did not have any mediating effects between visual frames and attitude. The findings of this study deepen our understanding of how visuals differ from textual framing in their effect on emotions and attitudes, and provide an opportunity to fill the gap in visual framing effect literature and the role of mediation in visual framing.
1 Introduction
The refugee crisis is arguably one of most urgent political issues facing European countries in recent history. Refugees fleeing war torn countries are flocking to Europe in mass waves, resulting in heated debates on how to handle this influx of people. Host communities are split between supporting and resenting the acceptance of refugees into their countries and villages (Yardley, 2016). Europe has witnessed an increase in anti-immigration sentiment, prompting a reaction by right wing groups to campaign against refugees in countries such as Sweden, Germany and the Netherlands (Scally, 2016; Deutsche Welle, 2016; The Economist, 2016; Sims, 2016). At the same time, the voice of counter movements in solidarity with refugees coming to Europe has also gained popularity (Brammall, 2015). Examples of such initiatives include the Refugees Welcome movement across Europe and the many active NGOs that provide help for refugees in their communities and abroad, such as Refugees Welcome accommodation share that started in Germany and spread across other European countries. Several organizations were established in the Netherlands and are now operating both inside and outside of the country such as Movement On The Ground, Boat Refugees (Boot Vluchteling) and Because We Carry.
Local and international media outlets have devoted extensive attention to this topic, covering news about refugees crossing borders, refugees’ living conditions in different European countries, and clashes between refugees and authorities. Media has multiple options in how to report the refugee crisis and what aspects they choose to emphasize, or neglect. Amidst all media coverage of the crisis, there was one powerful turning point, the release of the picture of Alan Kurdi, a three-year-old Syrian toddler whose body washed up on a Turkish shore after the boat he boarded with his family sunk. The image of his body lying face down in the sand was so powerful that it was widely circulated and triggered an international response: humanitarian aid and political debates on immigration policies. His image has been recognized as a symbol of the
refugee crisis that represents many children and people making the dangerous journey to Europe. But why was this picture so powerful? And what effects did that picture have on people for them to react the way they did on social media and in news articles?
Each image, sentence, or video constructs a message in a certain way, and presents a frame, a method of composing and delivering a specific message in a certain context. The way a message or an issue is presented has different effects on different people. Extensive research on framing effects has sufficiently demonstrated that frames in the media influence people’s emotions, attitudes and behaviors towards issues (e.g. Scheufele, & Iyengar 2012; Berinsky & Kinder, 2006; de Vreese and Boomgaarden, 2003; Brantner, Lobinger, & Wetzstein, 2011).
More specifically, studies showed that immigration has been covered in the media using different and possibly opposing frames (e.g. Van Gorp, 2005), and such frames affect how audiences feel about immigration. Emotions towards certain issues also have a mediating effect on how
people’s attitudes are formed (e.g. Lecheler, Bos & Vliegenthart, 2015). However, despite the multitude of studies focused on framing and framing effects, scholars argue that research is drifting away from the original meaning of framing and framing effects and confusing it with other theories (Scheufele & Iyengar, 2012). Therefore it is important to expand framing effect research to include non-verbal visuals, to return to what scholars argue is the traditional understanding of framing research (Scheufele & Iyengar, 2012 p.2). Yet, visual framing has
received less attention than textual framing, leaving a distinct gap in visual framing research, and
its effects on the public (Schuck & Feinholdt, 2015; Coleman, 2010). One of the main reasons visual framing has received less attention is the difficulty in identifying visual frames (Rodriguez & Dimitrova, 2011).
Although immigration has been studied before, little attention was paid to the visuals used to cover immigration or the refugee crisis. The refugee crisis is relatively recent, and there have been no studies examining the use of images in different frames in this crisis, and how they affect audiences. Yet, there has been discussion and public debate on the way media portrays refugees coming into European countries, and an even more heated debate on whether member states of the European Union should allow large numbers of refugees to enter, or instead close their borders. In 2015 alone, there were 43,093 asylum applications in the Netherlands
(VluchtelingenWerk Nederland). Although the Netherlands is a relatively small European country, it has witnessed an anti-refugee sentiment supported by right wing politician Geert Wilders (Sims, 2016). Because of this conflicted environment it is interesting to use the Netherlands as a case study to see how people react to different visual frames of refugees. Framing specifically is important to examine within this topic, because the Netherlands enjoys a high level of press freedom, ranking second worldwide (RSF, 2016). This makes the refugee crisis an even more critical issue to cover in the media given that Dutch media has greater freedom to choose how to report on it and what frames to use in their coverage. This in turn gives them power to influence the public opinion in the country regarding the refugee crisis.
Prior research has shown that there are indeed some radically opposing frames used to cover immigration topics, for example Van Gorp (2005) who used the news coverage of asylum seekers, refugees and illegal immigrants in Belgian newspapers as a case study to examine two issue specific frames: the intruder frame and the victim frame. This study derives from framing, framing effects and mediation theories, but deviates from the traditional textual frames and focuses on visual frames instead, building on the work of Van Gorp (2005). Besides looking at framing effects on emotions such as anger and empathy and people’s anti or pro refugee
attitudes, I will take a further step into examining how emotions mediate the impact of visual frames on attitudes. Thus, this study tackles the research question: What are the effects of visual
frames portraying refugees on people’s emotions and attitudes towards refugees in the
Netherlands?
2 Theoretical Frameworks
My theoretical framework is based on three main concepts: framing, framing effects and mediation. In this section, I explain how these concepts apply to visual framing. First, I define framing theory and connect it to visual framing in the context of the refugee crisis. Second, I introduce framing effects theory and address specifically effects on emotions and attitudes. Finally, I elaborate on how triggered emotions can influence the effect of visual frames on attitudes through mediation theory.
2.1 Visual Framing in the Media
To properly explain visual framing and later visual framing effects, it is important to understand the framing theory as a point of departure. Framing an issue in the media is less about what the topic is, and more about how the topic is reported. Framing in communication science is explained as different media reporting on the same topic in different ways by either highlighting or watering down certain aspects of the story, or emphasizing one reality over another to promote a particular perspective of the issue (e.g. Schuck & de Vreese, 2006; Entman, 1993). Media outlets that employ frames in ways that emphasize certain aspects over others possess the power to construct social realities (Schuck & de Vreese, 2006), and provide their own criteria for evaluating the issue in question (Nelson, Clawson, & Oxley, 1997, p.567-566). Media frames highlight the way a story is packaged to reflect both the representation and construction of a certain issue (e.g. Nelson & Kinder, 1996). Frames are used to paint a certain picture of reality
by media outlets and inject their agendas for public consumption. They are “perceptual schemes employed by communicators and audiences to organize and make sense of issues and events” and can influence audience’s perception of the topic in discussion (Greenwood & Jenkins, 2015, p.208). Frames can be even more critical when used in covering a sensitive or a controversial issue, because of their potential effect on the public, which I explore further in the next section.
However, framing is not limited to words and the textual representation of an issue. Images and videos can also be used to construct realities, emphasize certain aspects over others, and achieve the same goals of textual framing, perhaps with even more influence (Messaris & Abraham, 2001; Rodriguez & Dimitrova, 2011). Theoretically, there is an overlap between textual framing and visual framing. Leading scholars like Coleman (2010) and Trivundza (2006) suggest that framing theories are applicable to visual framing. Visual frames are argued to be especially powerful tools because they are “less intrusive than words and as such require less cognitive load” (Rodriguez & Dimitrova, 2011, p.50).
This paper focuses on the use of images in visual framing. Extant research has shown that the issue of immigration can be framed in different ways. As mentioned earlier, Van Gorp (2005) conducted textual content analysis of media coverage of asylum seekers in Belgium and
differentiated between two frames: the intruder frame and the victim frame. He categorized each frame by identifying certain elements used to portray asylum seekers as either victims or
intruders. Examples of the victim frame include the textual portrayal of the elderly, the child, or the poor. The intruder frame is the stereotypical ‘all strangers are intruders’ (Van Gorp, 2005), which considers refugees as “the others” who are seen as a threat to society and “one’s own cultural and economic achievements” (Van Gorp, 2005, p. 489). These elements can easily cross over to images as well. The setting of the image, the context, and the activity of the refugees
could reflect their vulnerability or possible aggressiveness. Media outlets have used images of children, women, elderly and vulnerable people to show refugees as victims. I argue that a visible portrayal of hardship, for example subjects crying, in danger, or without shelter, are elements used to emphasize a victim frame. On the other hand, images showing angry refugees trying to cross borders by force were used to show them as intruders and threats to society. Moreover, such a threat can also be portrayed through aggressive facial expressions. This study aims to use the two frames suggested by Van Gorp (2005) but in the form of visuals instead of text, and test people’s emotions particularly anger, fear, hope and empathy towards refugees, as well and their anti or pro refugee attitudes in the Netherlands.
2.2 Visual Framing Effects Theories
Scholars are in agreement that news frames have an effect on consumers of the news. Just as much as research has looked into different frames in the media, scholars have also studied the potential effects frames can have on people. Prior studies have looked at how framing in the media influences people’s emotions and attitudes. Extensive research on framing effects has sufficiently demonstrated that frames in the media influence people’s emotions (Berinsky & Kinder, 2006; Brantner et al. 2011; de Vreese & Boomgaarden, 2003) attitudes, and opinions
(e.g. Scheufele, & Iyengar 2012; Iyengar, 1994; Nelson, Oxley & Clawson, 1997; Aarøe, 2011). Different frames can have distinct effects depending on the frames used, but other factors, like people’s characteristics and prior attitudes influence the effectiveness of these frames (Scheufele, 2000).
As also explained by Scheufele and Iyengar, framing effects refer to the “behavioral or attitudinal outcomes” that occur not because of what is being communicated, but rather how the message is being delivered and the variations in how it is presented to the public (Scheufele &
Iyengar, 2012, p.1). When making even small changes to the way a massage is presented, there
will likely be visible changes in framing effects such as shifts in opinions or attitudes (Chong &
Druckman, 2007). A study by Levin and Gaeth (1988) shows that positive frames in the media evoke more positive responses, compared to negative frames. Photographs specifically are argued to be more effective tools than text because they give the audience the perception and feeling that they are as close to the truth as possible; they see what they believe to be reality through the picture (Messaris & Abraham, 2001; Rodriguez & Dimitrova, 2011). When putting visual frames and textual frames head to head, some scholars argue that visual framing wins
(Rodriguez & Dimitrova, 2011). For that reason, this study will look at visuals in the form of
images reflecting the refugee crisis in Europe.
2.2.1 Effects on emotions. Several studies have discussed the consequences of framing effects and suggested that framing influences people’s cognitive responses (e.g. Berinsky & Kinder, 2006; de Vreese & Boomgaarden, 2003; Nelson, Oxley & Clawson, 1997; Schuck & de Vreese, 2006). Emotions can be identified in relation to dimensions of valence and arousal when exposed to certain stimuli. This study focuses on valence, which refers to feelings of either pleasure or displeasure that can vary on the level of intensity as a response to the stimuli (Schuck & Feinholdt, 2015). For this study I categorize emotions into positive and negative. Negative
emotions include fear and anger, while positive emotions include empathy and hope. Studies
have looked at negative emotions extensively because of their importance when researching controversial topics. However, positive feelings have not received as much attention (Schuck & Feinholdt 2015) and therefore it is necessary to explore the role of positive emotions, as they
could have a motivating impact on people (Lecheler, Schuck & de Vreese, 2013;Schuck & de
Images are not very different from text when it comes to framing effects; in fact they could be more powerful. Visuals are perceived quickly and are able to activate emotions that will affect how recipients evaluate the entire message. Not only do photographs elicit emotion but they also convey human emotion more clearly than text (Brantner et al. 2011). Reporters and photojournalists use pictures to illustrate the most critical issues to capture the audience’s attention. Visuals are popular in war reporting especially; precisely because of the potential emotional effect photography has on viewers (Griffin, 2010). A content analysis study has shown that pictures of Palestinians who are depicted as victims or citizens are considered to be human-interest frame (Brantner et al., 2011). In the same study, the authors conducted an experiment that showed that human-interest frame has evoked empathy (Brantner et al. 2011). In another study, exposing participants to a victimization frame elicited positive emotions of hope and enthusiasm (Lecheler, Bos, & Vliegenthart, 2015). Similarly, the same relationship between images of Palestinian victims and people’s emotions could exist when depicting refugees as victims. As such, it is expected that people will feel empathy and hope when seeing images of refugees in a victim frame.
H1: Visual frames portraying refugees as victims will generate more empathy and hope compared to the intruder frame.
In contrast, the intruder frame is expected to generate the opposite response. The intruder frame is a package that represents potential threat and danger either to one’s self or society (Van Gorp, 2005). Being exposed to a frame that puts the audience in a situation where they feel threatened could elicit discrete emotions like anger and fear (Nabi, 2002). Fear specifically is elicited when one senses threat, or feels their mental and psychological situation is out of control. Consequently, such situations are expected to generate an attitude of avoidance. In a study on
immigration in the Netherlands, anger was found to be one of the prominent emotions elicited amongst participants exposed to immigration related frames (Verkuyten, 2004). On a different topic, climate change, participants experienced more anger when the issue was framed from a national security perspective, compared to other frames (Myers, Nisbet, Maibach, & Leiserowitz, 2012). A national security frame and an intruder frame both suggest danger and threat to people and society. Therefore, I expect that using frames depicting refugees as intruders or framing them as a potential threat to society will cause people to feel fear and anger as Nabi (2002) showed in her study. Thus, I hypothesize:
H2: Visual frames portraying refugees as intruders will generate more fear and anger compared to the victim frame.
2.2.2 Effects on attitudes. Various studies examine framing effects on people’s attitudes towards certain issues (Chong & Druckman, 2007; Nelson, Oxley & Clawson, 1997; Brantner et al. 2011). Studies that looked at textual framing of immigration suggest that people are more likely to develop anti-immigrant attitudes when immigrants are perceived to pose a real threat (Park, 2014). Visual frames have a powerful and lasting effect on the way viewers understand and comprehend certain issues (Nelson, Clawson, & Oxley, 1997). Photographs and images of conflict that are associated with extreme situations such as life and death like in times of war, for example, receive intense public attention, influencing the public’s perceptions of an issue and their attitudes toward related policies (Griffin, 2010). Both war and the refugee crisis are political crises that trigger emotions and attitudes. Therefore the frame of danger and threat is similar to that of an intruder frame.
A study by Boomgaarden and Vliegenthart (2007) found a positive correlation between the intensity of media coverage of immigration issues and support for anti-immigration parties.
Several studies examined the effects of frames in relation to immigration issues, finding that immigration frames have an effect on people’s opinions and attitudes towards immigrants (e.g., Brader, Valentino & Suhay, 2008; Igartua & Cheng, 2009). A previous study testing the effect of frames on tolerance has showed that people exposed to frames of danger and threat exhibited less tolerant attitudes (Nelson, Clawson, & Oxley, 1996). In a different study, participants who experienced emotions of sympathy, which is considered a positive emotion, tended to support immigration policies more than participants who experienced negative emotions like anger towards asylum seekers. They, in contrast, were less supportive of immigration policies
(Verkuyten, 2004). This paper will focus specifically on the extent of people’s tolerance towards welcoming refugees coming into the Netherlands. Based on this comparison I hypothesize: H3: People exposed to the intruder frame show less tolerant attitudes towards refugees compared to the victim frame.
2.3 Mediation Theory
Existing literature examines the relationship effect of framing on emotions, framing on attitudes, and emotions on attitudes, each relationship independent from the other. However there has not been much literature that studies emotions as mediators between frames and attitudes. To my knowledge, most mediation studies focused solely on textual frames and did not examine emotions as mediators specifically between visual frames and attitudes. For example one study did find that emotional reactions to positive and negative textual frames have a mediating effect
(Lecheler et al., 2013). Some studies suggest that news frames have an effect on people’s
emotions eliciting certain feelings, which in turn influence their attitudes (e.g. Kim & Cameron, 2011; Nabi 2002), depending on the content of the frame (Lecheler et al., 2015). Negative emotions like anger and fear, influence people to adopt more negative attitudes, while positive
emotions like empathy, and hope, solicit attitudes that are more positive (Lecheler et al., 2013).
A recent study showed that the effects of certain immigration frames were mediated by different emotions elicited during exposure. The study found that hope and anger mediated the effects of the emancipation frame, while the victimization frame was mediated by compassion and enthusiasm (Lecheler et al., 2015). In another study on attitudes towards immigrants in the Netherlands, anger was found to have a mediating effect on the relationship between national identification and support for immigration policy if refugees were seen as seeking asylum as a personal choice and not as economic or political refugees (Verkuyten, 2004). Similarly, anger was found to significantly mediate the effect of positive and negative frames on opinion
(Lecheler et al. 2013). Given the above argument that positive emotions were found to result in more positive attitudes and negative emotions to result in more negative attitudes, I expect that:
H4: Hope and empathy mediate the effect of the victim frame on attitudes H5: Anger and fear mediate the effect of the intruder frame on attitudes
Figure 1: Theoretical model
The theoretical model shows the relationships between all variables in the form of hypotheses as they were explained above. H1 and H2 are the direct effects of the victim frame on empathy and hope, and intruder frame on anger and fear respectively. H3 is the direct effect of both visual frames on attitude (tolerance). H4 and H4 are the indirect effect of the victim frame on tolerance via empathy and hope, and the indirect effect of the intruder frame on tolerance via anger and fear.
3 Methodology 3.1 Design and Procedure
To address all aspects of the research question, I conducted an experiment. Experiments are the most appropriate methodological approach to study framing effects, because they allow for testing short lived emotions that are experienced shortly after exposure to the stimuli
(Lecheler et al., 2015). The experiment was conducted during a two-week period from April 14th until April 28th of 2016.On average, the experiment took seven minutes for each subject to complete. The questionnaire was split into five parts: respondents were first asked basic
questions about their socio-demographic properties and background, as well as their opinion on a variety of issues and the extent to which they felt various issues were important. Following this, participants were randomly assigned to one of two visual frames, one depicting refugees as innocent victims and the other depicting refugees as intruders. After exposure to the stimuli, participants were asked to identify to what extent they felt different emotions, followed by statements to measure attitudes, with which they were asked to identify their level of agreement or disagreement. Manipulation checks were incorporated at the end of the survey, and
participants were debriefed after completion. 3.2 Participants
A total of 233 subjects participated in the experiment. After excluding respondents that did not complete the survey or answer all the necessary questions for this study, the final number of respondents was 135. Participants were recruited through snowball sampling via social media platforms where the survey was shared both directly with Dutch citizens and on student
Facebook groups and pages.People were also asked to share the survey with their family and Dutch friends. While it is important to acknowledge that the sample is a non-probability sample and therefore is not ideal, it is the most common sampling approach among Master students. All participants are Dutch citizens between the age of 18 and 69 (M= 36.8 years) have been living in the Netherlands for the past two years, and are familiar with the refugee crisis. Out of the total valid respondents 73 were female (54.1%) and 62 were male (45.9%). Randomization of age (intruder: [M=35.75, SD=12.58] versus victim: [M=37.94, SD=13.56], t= -.973, df=133, p=.332) and gender (intruder: [M=1.41, SD=.50] versus victim: [M=1.51, SD=.504]; t= -1.112, df=133,
3.3 Image Selection (Stimuli)
This study focused on two specific frames: victim and intruder. Two images that have been published in mainstream media as well as social media were selected based on a set of criteria. The first round of image selection was based on three overall criteria. Images were specifically chosen to represent an episodic human-interests frame (Semetko & Valkenburg, 2000). This means that images had specific characters at which receivers could direct their emotional reactions (Aarøe, 2011), and also help put faces to the issue, in order to bring an emotional angle to the presentation of the refugee crisis (Semetko & Valkenburg, 2000). Second, images did not have any captions or texts to ensure that framing effects were due to the image and not the text. Lastly, the pictures were taken from the past year (2015) and reflect the latest status of the refugee crisis in Europe. To ensure that the images depicted the two intended
frames, victim and intruder, I took additional aspects into consideration. Certain visual indicators are considered when choosing images to present a frame. An example of indicators that were previously considered in a visual framing study included the roles of the subject in the pictures
(civilians versus combatants), the main activity depicted in the picture, the age and gender of the
subjects, and whether injury or harm were depicted (Brantner et al., 2011). For the victim frame, those indications were well thought out. The chosen image (figure 2) portrays a father and a child at a refugee camp. The portrayal of a child is one aspect of the victim frame. The other aspect is the visible expression of emotion by the father who is seen crying as his son comforts him. The scene takes place in a refugee camp, and the father is holding a piece of bread that has been presented to him at the camp. This paints an experience of hardship, which I earlier argued is another aspect of the victim frame. As for the intruder frame, also like mentioned earlier, the subjects in the image (figure 3) show clear aggressive facial expressions as they try to cross
borders by force. Force is reflected in the long metal poles they are using to break the fence, which can be interpreted as a danger or threat. I conducted a pretest with a few respondents and the results indicated that the two frames were indeed perceived as intended.
Figure 2: The victim frame; a child trying to comfort his crying father at a refugee camp.
3.4 Measurements
3.4.1 Emotions. The type of emotion to be measured is emotional valence, the level and extent of participants’ feelings after being exposed to the stimuli (Schuck & Feinholdt, 2015). To measure emotions immediately after stimuli, I used a self-report approach where participants responded to nine emotions using a 7-point-Likert-scale ranging from 1 (not at all) to 7 (very much) as previously used in other experimental studies (e.g. Lecheler et al., 2015; Verkuyten, 2004; Kim & Cameron, 2011). The scale tested for the following nine emotions, which were used in the studies mentioned above: satisfaction, compassion, enthusiasm, hope, anger, fear, grief, concern and empathy. This study is focused on the hope, empathy, anger and fear emotions.
3.4.2 Attitudes. Participants were asked to state whether they agree or disagree with five statements designed to test their tolerance towards welcoming refugees to the Netherlands using a 7-point-scale from 1 (strongly disagree) to 7 (strongly agree) as used in other studies (e.g. Kim
& Cameron, 2011). The Cronbach's alpha for the tolerance scale suggests it was very reliable (5
items; α = .759). The statements that were used to measure tolerance are: 1. Cultural life in Dutch communities is disrupted when foreigners come to live here
2. It's a good thing for the Dutch economy if people from other countries come to live here 3. The Netherlands has become a better place to live in since the refugees came to live here
4. Cultural life is disrupted in Dutch communities when foreigners come to live here 5. The Netherlands is still able to take in more refugees
6. There should be stricter rules for foreigners who want to come to live and work in the Netherlands.
3.4.3 Controlled variables. The key independent variable is the frame, which tracks
whether a respondent was exposed to the intruder frame (0) or the victim frame (1). In addition to this, I also measured various control variables that I expected might affect the resuls: prior attitudes, age and gender. Literature suggests that existing or prior attitudes towards certain issues determine how people process information and react to frames, influencing their formation of opinion and attitudes (Schuck & de Vreese, 2009). I measured for prior attitudes on refugees coming into the Netherlands, and several other topics to hide the main purpose of the
measurement. Participants were asked to state on a 7-point-scale whether they agree or disagree
with the statements (translated from Dutch): 1. The Netherlands should take in more refugees. 2.
The majority of refugees who enter the Netherlands are dangerous. Moreover, I also measured
participants’ age and gender - female (1) or male (2).
3.4.4 Manipulation checks. To ensure proper reception of the visual frames participants were asked to rate whether they agree or disagree with five statements on a 7-point-scale ranged from 1 (strongly disagree) to 7 (strongly agree). The statements were identical for both stimuli, and described different aspects of the picture such as facial expressions; emotions visible in the picture, perception of potential threat. Participants were expected to have opposite responses to each statement. Overall results show that people responded well to manipulation checks, with significant differences between the two groups in how they perceived the visuals. When asked whether refugees in the picture showed emotions, there was high significance in difference between respondents in the victim frame (M=6.58, SD=0.99), and respondents in the intruder frame (M=5.91, SD=1.38; t= -3.251, df=121.440, p<.01). Participants were also asked to rate whether they thought the subjects in the picture showed aggressive facial expressions to which there was significantly more agreement in intruder frame (M=4.47, SD=1.63) than in the victim
frame (M=1.31, SD=0.68, t=14.696, df=89.730, p<.001). Participants exposed to the intruder frame also had a significantly higher level of agreement that the subjects in the picture were violating the law (M=3.54, SD=1.82) compared to participants exposed to the victim frame (M=1.37, SD=0.89, t=8.851, df=97.444, p<.001). Responding to whether refugees depicted in the picture showed weakness, results show higher agreement in the victim frame (M=2.36, SD=1.54) versus the intruder frame (M=2.36, SD=1.54), however it is not statistically significant (t=-1.635,
df = 123.716, p=.104). Finally participants were asked whether they thought the people in the
picture could pose danger to the community in which they live. Results showed a significant difference between the two conditions (t=3.702, df=120.951, p<.001), with participants in the intruder frame agreeing more (M=2.31, SD=1.48) compared to the victim frame (M=1.49,
SD=1.05). Findings confirm that different aspects were indeed perceived oppositely between the
frames as expected.
4 Results 4.1 Felt Emotions
Figure 4 shows the difference in levels between the four main emotions in each
condition. As previously stated, the hypotheses were that participants who were exposed to the visual frame would experience higher level of empathy and hope compared to the intruder frame (H1), while participants exposed to the intruder frame would experience more fear and anger compared to the victim frame (H2).
Figure 4: Comparison of levels of hope, empathy, anger, and fear felt between the two visual
frames: Victim and Intruder.
Note: Numbers on bars represent the mean for each emotion in the two conditions.
To compare felt emotions for each condition (victim and intruder frames), I conducted independent T-Tests on each of the four main emotions: hope, empathy, anger and fear. Results show that participants exposed to the victim frame indeed experienced significantly more empathy (M=5.88, SD=1.35) than those exposed to the intruder frame (M=5.03, SD=1.57; t= -3.366, df=133, p<.01). Similarly, results show that feelings of hope are significantly higher for participants in the victim frame (M=3.19, SD=1.80) than in the intruder frame (M=2.34,
SD=1.51; t= -2.990, df=133, p<.01), which confirms H1. Second, I examined anger and fear in
the two conditions. Contrary to my expectation, feelings of anger were slightly higher in the victim frame (M= 4.28, SD=1.76, t= -.732, df=133, p=.466) compared to the intruder frame (M=4.06, SD=1.81), though not statistically significant. As for fear, although results suggest that
fear is indeed felt slightly more in the intruder frame (M=2.82, SD=1.76) compared to the victim frame (M=2.57, SD=1.73), it is statistically insignificant (t= -.854, df=133, p=.395), thus H2 is rejected.
4.2 Tolerance for Refugees Coming to Netherlands
Participants’ answers were tested for their level of tolerance towards having refugees come to live in the Netherlands under each framing condition. Attitudes were measured by participants’ indications of whether they agree or disagree with five different statements.
Looking at all five statements1, overall participants in the intruder frame showed slightly more tolerance (M=25.32, SD=5.15) compared to participants in the victim frame (M=24.12,
SD=6.58), which oppose the hypothesis. The difference between their levels of tolerance is insignificant (t= 1.180, df=132, p=.240), but these findings contradict the expectations and therefore refute H3.
As these findings run counter to the expectations, I examined the separate statements. The first statement is “It's a good thing for the Dutch economy if people from other countries
come to live here”. Unexpectedly, participants in the intruder frame had higher level of
agreement (M=5.15, SD=1.39) compared to those in the victim frame (M=4.90, SD=1.43).
However the difference between two frames is not significant (t= 1.04, df=133, p=.300). The
second statement is “cultural life in Dutch communities is disrupted when foreigners come to live
here.” This statement is a negative one, therefore the lower the level of agreement with it, the
more tolerant the participant is considered in this case. Participants in the intruder condition
1 The sum of attitude’s mean was not divided by five items, which is why it has a high value.
(M=2.06, SD=1.28) are again unexpectedly slightly more tolerant than participants in the victim
(M=2.55, SD=1.65) condition.2The second statement is insignificant, though barely (t= -1.906,
df=122.645 p=.059). The third statement “The Netherlands has become a better place to live in
since the refugees came to live here,” again show no significant difference between the two
frames (t= 0.65, df=133 p=.948), with a slightly higher mean in the intruder frame (M=3.90,
SD=1.48 versus M=3.88, SD=1.45). Similarly, participants who were exposed to the intruder frame agree slightly more than participants in the victim frame to the statement “The Netherlands
is still able to take in more refugees” (Victim frame M=5.24, SD=2.02 versus M=5.41, SD=1.90),
however there was no significant difference (t= 0.512, df=133 p=.610). Finally the last statement is “There should be stricter rules for foreigners who want to come to live and work in the
Netherlands.” Participants exposed to the intruder frame (M=3.07, SD=1.75) showed less
agreement that there should be stricter rules regulating work and living policies for refugees in the Netherlands than participants exposed to the victim frame (M=3.27, SD= 2.00). Consistent with the previous four statements, despite differences in means the statistical difference is not significant (t= -0.603, df=133 p=.547).
Even though randomizing succeeded as mentioned earlier, the findings of framing effects on people’s tolerance towards refugees were the opposite of expectations. Moreover, even after analyzing the results for each statement separately, the results were still contrary to expectation. Therefore I controlled for other variables to further explore whether participants’ attitudes could
2 The results for the second statement are missing one participant for the victim frame who failed
to answer the question. I did not delete his data for attitude measurement, because the missing data from one statement does not obscure the overall results.
be affected by other factors like prior attitudes, age and gender. Prior attitudes were measured by conducting an Independent sample T-test and the results show that participants who were
assigned to the intruder frame had a slightly higher level of tolerance towards refugees before the stimuli (M=11.66, SD=2.42) compared to participants who were assigned to the victim frame
(M=11.43, SD= 2.77) however the difference is not significant (t= .511, df=133 p=.610).
A multiple regression was used to test if participants’ prior attitudes, visual frames, age, and gender (coded female=1, male=2) significantly predict participants’ tolerance, after the stimuli. This multiple regression accounted for 56.4% of the variability (R2: .564). The analysis was found to be statistically significant F(4,129) = 41.760, p < .000, indicating the four
variables are good predictors of participants’ tolerance towards refugees. It was found that both
the victim frame and gender were not significant predictors of tolerance towards attitudes, (β =
-.877, p=.206) and (β = -.072, p=.917) respectively. On the other hand both age (β = .059, p<.05)
and prior attitudes (β = 1.660, p<.000) were found to be significant predictors for participants’
tolerance. However even after controlling for prior attitudes, tolerance was not affected and remained insignificant between frames.
4.3 Mediation Analysis
Finally, H4 and H5 expected that the impact of the frames on attitudes would be mediated through the four emotions. Specifically that empathy and hope would positively mediate the effect of the victim frame on tolerance, and that anger and fear would negatively mediate the effect of the intruder frame of tolerance. Scholars argue that bootstrap confidence intervals are the most powerful method to properly assess significance of mediation or indirect effects (Preacher, Rucker, & Hayes, 2007; Lecheler et al., 2013) compared to the ‘causal-steps
prerequisite conditions for mediation testing that could compromise results of possible indirect effects between variables. Same authors also criticized The Sobel test approach, another method to test mediation, arguing it is only suitable for large samples. Therefore I applied the bootstrap method introduced by Preacher and Hayes (2004; 2008).
Figure 5: Multiple mediation model for the indirect effect of the victim and intruder
frames on tolerance via empathy, hope, anger, and fear.
Note: The direct effect of the victim frame (X1) on tolerance (Y) is illustrated in c’1, while c’2 is
the direct effect of the intruder frame (X2) on tolerance (Y). The indirect effect of both visual frames on tolerance via the mediators M1 (empathy), M2 (hope), M3 (anger), M4 (fear) is illustrated as a1 b1, a2 b2, etc. The total effect of visual frames on attitude is the sum of both the direct effects and the specific indirect effects (a and b arrows).
To conduct the analysis, I used Preacher and Hayes’ (2008) INDIRECT Macro in SPSS. Results revealed that significant indirect effect of the victim frame on tolerance mediated via
empathy (b = 1.16, SE = .43) (95% bca CI: [.440; 2.144]) compared to the intruder frame. Whereas fear (b = .257, SE = .35) (95% bca CI: [-.384;1.024]), anger (b = .120, SE = .20) (95% bca CI: [-.180; .703]), and hope (b = -.23, SE = .26) (95% bca CI: [-.844; .208]) were found to have no significant mediating effects of visual frames on participant’s tolerance towards
refugees. The victim frame has a positive effect on empathy (b = .83, SE = .25, p < .01), which in turn contributed to higher levels of tolerance towards refugees coming to the Netherlands (b = 1.39, SE = .34, p < .001) compared to the intruder frame. Although the victim frame had a positive effect on hope (b = .89, SE = .29, p < .01), hope had an insignificant impact on tolerance among participants (b = −.26, SE = .29, p=. 37). Anger (b = .23, SE = .31, p=.46) and fear (b = −.23, SE = .30, p=.44) were both not significantly affected by the visual frames, but while anger had no significant impact on tolerance (though close) (b = .53, SE = .29, p=.07), fear did (b = −1.11, SE = .28, p < .001). The explained variance of the full mediation model (R2 = .01) is not significant F(1, 132) = 1.39, p = .240.These findings partially confirm H4 and completely refute H5.
5 Discussion and Conclusion
Many scholars have studied framing effects on emotions and attitudes (e.g. Brantner et al. 2011; de Vreese & Boomgaarden, 2003; Scheufele, & Iyengar 2012; Nelson, Oxley & Clawson, 1997; Aarøe, 2011). Visual framing--images specifically--are argued to have a stronger effect than text because the audience perceives images as closer to reality (Messaris & Abraham, 2001;
Rodriguez & Dimitrova, 2011). In this study I chose the current refugee crisis as a case study to
look at the effects of images of refugees on people’s emotions and attitudes in the Netherlands. I conducted an experiment that exposed participants to one of two frames, victim or intruder, based on Van Gorp’s study (2005). The concentration of my analysis was on the emotions of
hope, empathy, fear and anger and the attitude of tolerance towards refugees. Based on previous literature I hypothesized that participants assigned to the victim frame would experience higher levels of hope and empathy compared to participants assigned to the intruder frame, whereas anger and fear would be experienced more in the intruder frame compared to the victim frame.
Some of the study’s findings accord with literature that frames evoke emotions (e.g.
Aarøe, 2011; Berinsky & Kinder, 2006; Nelson, Oxley & Clawson, 1997). More specifically
findings regarding hope and empathy were in line with literature, which found that the victim frame depicting refugees as vulnerable people in a helpless situation elicited significantly higher levels of hope (Lecheler et al., 2015) compared to the intruder frame. Similarly, empathy was significantly higher in the victim frame compared to the intruder frame (H1). This finding expands on the finding of previous literature, which suggests that human-interest frames elicit
empathy (Brantner et al. 2011) by connection of this emotion to victim frame specifically. Other
results show findings that are in opposition to previous literature, which suggest fear and anger are elicited by frames representing potential threat or danger (Nabi 1999 & 2002; Verkuyten, 2004). Contrary to expectation, fear was not significant, and anger was neither prominent, nor experienced in connection with threat and danger (intruder frame) (H2). In fact quite the
opposite, anger was felt slightly higher--but not significantly so--in the victim frame compared to the intruder frame. One explanation could be that the victim frame elicited feelings of anger not necessarily aimed at the refugees but at the situation refugees are facing, or perhaps anger at the lack of emergency response to help vulnerable refugees who are perceived as victims.
I expected that people exposed to the victim frame would be more tolerant towards accepting refugees in the Netherlands compared to people exposed to the intruder frame. The findings of this study are in contrast to previous literature regarding framing effects on attitudes. Despite the
argument that I and others proposed, that visual frames could be stronger than textual frames in shaping people’s attitudes on certain issues, my study found no significant effect on participant’s attitudes towards refugees from visual frames. Griffin (2010) argued that images associated with extreme life and death situations influence people’s attitudes towards policies and issues, though my findings suggest otherwise. Additionally my study refutes arguments that immigration related frames have an effect on people’s attitudes towards immigrants (e.g., Brader, Valenntino & Suhay, 2008; Igartua & Cheng, 2009), in this study towards refugees. Unexpectedly, there was no positive correlation between depictions of refugees as a threat and anti-refugee sentiments. Tolerance was in fact higher, though not significant, in the intruder frame where refugees were presented as threat and posing danger, compared to the victim frame, contrary to what Park (2014), Nelson et al. (1996) and Verkuyten (2004).)suggest. Perhaps measuring subject reaction to a single image does not account for long-term attitudes, which might have been deeply rooted prior to the experiment and are not easily affected by one image. I found that participants who were assigned to the victim frame had less tolerance for refugees prior to the stimuli, while participants who were randomly assigned to the intruder frame had higher tolerance. Participants’ attitudes remained consistent even after being exposed to the stimuli, where
participants in the intruder frame still had more tolerance than participants in the victim frame as
mentioned above. This could be explained by the motivated reasoning theory, which suggests
that people who have strong prior attitudes are least receptive to new information and have the “tendency to evaluate incoming information to support preconception and to devalue contrary evidence” (Lodge & Taber, 2000, p. 184).
Finally, I expected hope and empathy to mediate the effect of the victim frame on participant’s tolerance level, and anger and fear to mediate the effect of the intruder frame on
tolerance. Although some scholars like Baron and Kenny argue that there has to be a direct effect between the independent variable (X) and the independent variable (Y) for there to be mediation, Hayes disagreed and argued there could be indirect effects through mediators (Hayes, 2009). Therefore, despite not finding a significant main effect of visual frames on people’s levels of tolerance, I conducted mediation analysis to see whether the four emotions play mediating roles like literature suggests. Findings show that only some emotions had a mediating effect between visuals and attitudes. This study found that empathy, a positive attitude, significantly mediated the effects of the victim frame on people’s level of tolerance, which is somewhat in line with a
previous study that found the victimization frame was mediated by compassion and enthusiasm,
also positive attitudes (Lecheler et al., 2015). However contrary to the suggestion of Lecheler et al. (2013) that positive emotions elicit positive attitudes, I did not find hope to have a positive influence on people’s attitudes. I also found that anger was neither affected by the visual frame nor did it have a mediating effect. This finding opposes the findings of both Lecheler et al. (2013) that anger has a direct effect on attitude, and in return contributes to the framing effects on attitudes, and Verkuyten’s (2004) that anger mediated the effect of frames and influenced participants to adopt anti immigration attitudes. This study found that anger was neither affected by the visual frame nor did it have a mediating effect, contrary to the arguments of these studies.
Verkuyten (2004) also argued that positive emotions generated positive attitudes towards immigrants, and negative emotions generated more negative attitudes towards immigrants, however the findings of this study show different results.
6 Limitations and Future Research
This study focuses on visual frames only because they deserve much needed attention in the academic field. However, future researchers should consider looking at visuals that are
accompanied with texts or captions, and test whether the effects are similar or different. It is argued that it might be easier for viewers to make more sense of visuals which are accompanied with texts that represent a clear frame, because visuals alone are not as explicit in explaining cause-and-effect relationships (Rodriguez & Dimitrova, 2011).
Visuals in this study were limited to one image per frame (two images in total). Although it is common to use single exposure in framing experiments (e.g., Druckman & Nelson, 2003), perhaps future studies should consider multiple exposures. Given that results found no effect of visual frames on attitudes, one image might not have an effect, and instead subjects would need to be repeatedly exposed to multiple images for their attitudes to be influenced. In real life, news coverage is consistently presenting frames and visuals as opposed to the single stimulus of this experiment.
An interesting addition would be to follow one study’s argument that positive frames mobilize people to support policies (Schuck & de Vreese, 2009), and research what visual frames are considered positive in crises like the refugee issue, and where positive visual frames indeed lead to mobilization of people to support positive policies towards issues. Furthermore, one could take a deeper look at media outlets’ choices of visual frames, and whether they purposely choose negative or positive visual frames to rally their audiences either in support of or against certain policies based on their affiliation. More specifically it would be interesting to look at the types and variety of frames used within one media outlet, and whether media outlets use either the victim or intruder frame, or a mix of both when covering issues like the refugee crisis.
In terms of methodology, the choice of research design to conduct this study has the disadvantage of having a low external validity and generalizability, however it is argued to be the best method to test effects as mentioned above. The sample is also a non-probability sample
accessed via snowball sampling among university students, which is not the ideal sample for representative results. However, the use of student sampling is common and is argued to be a generalizable because the behavior of students is not significantly different from that of a more diverse sample (Druckman, 2001). Furthermore, my sample goes beyond university students and expands to a larger population. One limitation important to note, however, is that my sample had a generally positive attitude towards refugees before being exposed to the experiment, and future research should include more variety in prior attitudes. Future research could expand the sample size over a longer period and target the larger Dutch population to have a more representative sample. Future research could also further expand to the larger European population and study visual framing effects on people in other countries also experiencing the refugee crisis, like Germany and Sweden.
Another limitation is that emotions in this study are measured using a self-report
approach. This means that results on emotions felt in each frame depend on participants’ ability to identify their feelings at the moment. Despite the fact the people might not be entirely sure how they are feeling, results from my study have shown significance in some emotions like hope and empathy, which suggests that most participants experience similar feelings. However, a way to improve this study and build on it in the future is to use more scientific measurement
approaches like physiological indicators that are able to more accurately capture emotional and cognitive responses to stimuli (e.g. Daignault, Soroka, & Giasson, 2013; Codispoti, Ferrari, & Bradley, 2006). Also, as mentioned in the discussion section, emotions were measured only once, immediately after participants were exposed to the stimulus, which means emotions were most likely short lived and activated by the image they saw. Building on this study, future scholars can investigate long-term effects on emotions by conducting a longitudinal study.
Additionally, this study had two groups that were exposed to two conditions. Future researchers could also add a control group that would allow for comparisons between frames that affect emotions and attitudes, and those that do not.
Finally, choosing the refugee crisis as the topic for this study has advantages and disadvantages. Refugees and immigration related issues have been studied before, specifically the victim and intruder frames, however focusing on visuals brought a new dimension to the literature on visual framing in immigration related issues. Additionally, the topic is very relevant politically and socially. However, because it is a popular topic, people’s attitudes towards this issue might have already been established based on exposure to prior media coverage, which might be one reason I found no significant effect on attitudes. Future work should consider using the victim and intruder frames in relation to other issues, and test whether salient or non-salient issues are more appropriate to better understand the effects of visual frames.
This study accords with literature that suggests frames have effects on emotions. More importantly this study confirms that framing effect theories on emotions are indeed be applicable to visual frames. At the same time I highlighted faults with theories that suggest framing has effect on attitudes, and more specifically that framing danger and threat generates negative attitudes towards refugees. Despite not finding results that support this argument in visual
framing, this research provides a platform for new studies to fill the gap in visual framing effects literature, especially in coverage of ongoing issues like the refugee crisis. I suggest that future research replicate this experiment to measure different emotions and frames, to reach more concrete conclusions on how visual framing effects differ from textual framing effects on different emotions and attitudes.
References
Aarøe, L. (2011). Investigating frame strength: The case of episodic and thematic frames. Political Communication, 28(2), 207-226.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of
personality and social psychology, 51(6), 1173.
Berinsky, A. J., & Kinder, D. R. (2006). Making sense of issues through media frames: Understanding the Kosovo crisis. Journal of Politics, 68(3), 640-656.
Brader, T., Valentino, N. A., & Suhay, E. (2008). What triggers public opposition to
immigration? Anxiety, group cues, and immigration threat. American Journal of Political
Science, 52, 959-978.
Brammall, S. (2015, October 9). How Europe's charities are adapting and innovating to help refugees. The Guardian. Retrieved from http://www.theguardian.com/voluntary-sector-network/2015/oct/09/facebook-refugees-charities-europe-collaboration-ngo.
Brantner, C., Lobinger, K., & Wetzstein, I. (2011). Effects of visual framing on emotional responses and evaluations of news stories about the Gaza conflict 2009. Journalism &
Mass Communication Quarterly, 88(3), 523-540.
Boomgaarden, H.G., & Vliegenthart, R. (2007). Explaining the rise of anti-immigrant parties: The role of news media content. Electoral Studies, 26, 404–417.
Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science, 10, 103-126.
Codispoti, M., Ferrari, V., & Bradley, M. M. (2006). Repetitive picture processing:Autonomic and cortical correlates. Brain Research, 1068(1), 213–220.
Coleman, R. (2010). Framing the pictures in our heads. In P. D'angelo & J. A. Kuypers (Eds.)
Doing News Framing Analysis: Empirical and Theoretical Perspectives (pp. 233-261)
New York, NY: Routledge.
Daignault, P., Soroka, S., & Giasson, T. (2013). The perception of political advertising during an election campaign: A measure of cognitive and emotional effects. Canadian journal of
Communication, 38(2), 167-186.
Deutsche Welle. (2016, February 22). In Germany, pressure mounts to handle right-wing violence against refugees 22.02.2016. Deutsche Welle. Retrieved from
http://www.dw.com/en/in-germany-pressure-mounts-to-handle-right-wing-violence-against-refugees/a-19065433
De Vreese, C., & Boomgaarden, H. (2003). Valenced news frames and public support for the EU. Communications, 28(4), 361-381.
Druckman, J. N. 2001. On the limits of framing effects: Who can frame? Journal of Politics, 63, 1041–1066.
Druckman, J. N., & Nelson, K. R. (2003). Framing and deliberation: How citizens' conversations limit elite influence. American Journal of Political Science, 47(4), 729-745.
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of
Communication, 43(4), 51-58.
Greenwood, K., & Jenkins, J. (2015). Visual Framing of the Syrian Conflict in News and Public Affairs Magazines. Journalism Studies, 16(2), 207-227.
Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication monographs, 76(4), 408-420.
Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, moderation, and conditional process modeling. Retrieved from http://www.afhayes.com/ public/process2012.pdf.
Igartua, J. J., & Cheng, L. (2009). Moderating effect of group cue while processing news on immigration: Is the framing effect a heuristic process? Journal of Communication, 59, 726-749.
Iyengar, S. (1994). Is anyone responsible?: How television frames political issues. Chicago, IL: University of Chicago Press.
Kim, H. J., & Cameron, G. T. (2011). Emotions matter in crisis: The role of anger and sadness in the publics' response to crisis news framing and corporate crisis
response. Communication Research, 0093650210385813.
Lecheler, S., Schuck, A. R., & de Vreese, C. H. (2013). Dealing with feelings: Positive and negative discrete emotions as mediators of news framing effects. The European Journal
of Communication Research, 38 (2), 189 - 209.
Lecheler, S., Bos, L., & Vliegenthart, R. (2015). The Mediating Role of Emotions News Framing Effects on Opinions About Immigration. Journalism & Mass Communication Quarterly,
92(4), 812-838.
Levin, I. P., & Gaeth, G. J. (1988). How consumers are affected by the framing of attribute information before and after consuming the product. Journal of consumer research, 15(3) 374-378.
Lodge, M., & Taber, C. (2000). Three steps toward a theory of motivated political reasoning. In A. Lupia, M.D. McCubbins & S.L. Popkin (Eds.) Elements of reason: Cognition, choice,
and the bounds of rationality (pp. 183-213). Cambridge, UK: Cambridge University
Press.
Messaris, P., & Abraham, L. (2001). The role of images in framing news stories. In S. Reese et al. (eds)Framing public life: Perspectives on media and our understanding of the social world, (pp. 215-226). Mahwah, NJ: Lawrence Erlbaum.
Myers, T. A., Nisbet, M. C., Maibach, E. W., & Leiserowitz, A. A. (2012). A public health frame arouses hopeful emotions about climate change. Climatic Change, 113(3-4), 1105-1112. Nabi, R. L. (2002). Anger, fear, uncertainty, and attitudes: A test of the cognitive-functional
model. Communication Monographs, 69(3), 204-216.
Nelson, T. E., & Kinder, D. R. (1996). Issue frames and group-centrism in American public opinion. Journal of Politics, 58, 1055-1078.
Nelson, T. E., Clawson, R. A., & Oxley, Z. M. (1997). Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review, 91(03), 567-583.
Nelson, T. E., Oxley, Z. M., & Clawson, R. A. (1997). Toward a psychology of framing effects.
Political behavior, 19(3), 221-246.
Park, K. (2014). Foreigners or multicultural citizens? Press media's construction of immigrants in South Korea. Ethnic and racial studies, 37(9), 1565-1586.
Preacher, K., & Hayes, A. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments Computers, 36, 717– 731.
Preacher, K., & Hayes, A. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation
hypotheses: Theory, methods, and prescriptions. Multivariate behavioral research, 42(1), 185-227.
Rodriguez, L. & Dimitrova, D. V. (2011). The levels of visual framing. Journal of Visual
Literacy, 30(1), 48-65.
RSF. (2016, April 26). Classement de la liberté de la presse 2016. Reporters Without Borders. Retrieved April 27, 2016 from https://rsf.org/en/ranking_table.
Scally, D. (2016, January 31). Swedish police warn of upsurge in far-right violence. The Irish
Times. Retrieved from
http://www.irishtimes.com/news/world/europe/swedish-police-warn-of-upsurge-in-far-right-violence-1.2517246
Scheufele, D. A. (2000). Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication and Society, 3(2–3), 297–316.
Scheufele, D. A., & Iyengar, S. (2012). The state of framing research: A call for new directions. In K. Kenski & K.H. Jamieson (Eds.) The Oxford Handbook of Political
Communication Theories (pp. 1-26). New York, NY: Oxford University Press.
Schuck, A. R., & De Vreese, C. H. (2006). Between risk and opportunity news framing and its effects on public support for EU enlargement. European Journal of Communication,
Schuck, A. R., & De Vreese, C. H. (2009). Reversed mobilization in referendum campaigns how positive news framing can mobilize the skeptics. The International Journal of
Press/Politics, 14(1), 40-66.
Schuck, A. R., & Feinholdt, A. (2015). News Framing Effects and Emotions. In R. A. Scott & S.M. Kosslyn (Eds.) Emerging Trends in the Social and Behavioral Sciences: An
Interdisciplinary, Searchable, and Linkable Resource. Hoboken, N.J.: John Wiley &
Sons.
Semetko, H. A., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news. Journal of communication, 50(2), 93-109.
Sims, A. (2016, January 23). Far-right Dutch politician, Geert Wilders, says male refugees must be kept in ‘asylum camps’ to stop ‘sexual jihad’. The Independent. Retrieved from http://www.independent.co.uk/news/world/europe/far-right-dutch-politician-geert-wilders-says-male-refugees-must-be-kept-in-asylum-camps-to-stop-a6828891.html The Economist. (2016, February 27). Radikale Rechte. The Economist. Retrieved from
http://www.economist.com/news/europe/21693636-after-falling-years-xenophobic-extremism-germany-rise-radikale-rechte
Trivundža, T. I. (2006). Many voices, one picture: photographic coverage of foreign news in Slovenian daily press (1980, 2004). Javnost-The Public, 13(2), 21-40.
Van Gorp, B. (2005). Where is the frame? Victims and intruders in the Belgian press coverage of the asylum issue. European Journal of Communication, 20 (4), 484 - 507.
Verkuyten, M. (2004). Emotional reactions to and support for immigrant policies: Attributed responsibilities to categories of asylum seekers. Social Justice Research, 17(3), 293-314.
VluchtelingenWerk Nederland. (n.d.). Bescherming in Nederland. Retrieved April 29, 2016 from http://www.vluchtelingenwerk.nl/feiten-cijfers/cijfers/bescherming-nederland
Yardley, J. (2016, January 13). Sexual Attacks Widen Divisions in European Migrant Crisis. The
New York Times. Retrieved from
